NH9.2 | New data, methods and opportunities to explore natural hazards, societal vulnerabilities and disasters in an interconnected world
EDI

Increasing effects of climate change, urbanization, and increased interconnectedness between ecological, physical, human, and technological systems pose major challenges to disaster risk management in a globalised world. Economic losses from natural hazards and climate change are still increasing, and the recent series of catastrophic events across the world together with the COVID-19 crisis and ongoing conflicts have manifested the need to shift from single-hazard and sectoral approaches to new and innovative ways of assessing and managing risks across sectors, borders and scales based on a multi-hazard and systemic risk lens.

Addressing the above challenges, this session aims to gather the latest research, empirical studies, and observation data that are useful for understanding and assessing the complex interplay between multiple natural hazards and social vulnerabilities to: (i) identify persistent gaps, (ii) propose potential ways forward, and (iii) inform resilience building strategies in the context of global change.

Co-organized by GI6/HS13
Convener: Johanna MårdECSECS | Co-conveners: Korbinian Breinl, Michael Hagenlocher, Giuliano Di Baldassarre
Orals
| Mon, 24 Apr, 16:15–18:00 (CEST)
 
Room 1.15/16
Posters on site
| Attendance Mon, 24 Apr, 10:45–12:30 (CEST)
 
Hall X4
Posters virtual
| Attendance Mon, 24 Apr, 10:45–12:30 (CEST)
 
vHall NH
Orals |
Mon, 16:15
Mon, 10:45
Mon, 10:45

Orals: Mon, 24 Apr | Room 1.15/16

Chairpersons: Johanna Mård, Korbinian Breinl, Giuliano Di Baldassarre
16:15–16:20
16:20–16:30
|
EGU23-286
|
NH9.2
|
ECS
|
On-site presentation
Jan Sodoge, Christian Kuhlicke, Miguel Mahecha, and Mariana de Brito

Socio-economic drought impacts often occur concomitantly across multiple sectors, leading to more severe consequences than if they affected single sectors. Improved management of such disasters requires cross-sectoral impact assessments and analyses. As such, analyzing how regions are affected by multiple impacts can provide crucial information for mitigating their consequences. Here, we characterize the multivariate distributions of socio-economic drought impacts. Our aim is to understand patterns by which diverse drought impacts co-occur. We introduce the concept of drought impact profiles, which describe characteristic distributions of co-occurring impacts. To this end, we use a unique spatio-temporal dataset generated with text mining and machine learning applied to newspaper articles. This dataset describes reported socio-economic drought impacts along seven categories (agriculture, forestry, fires,  social, aquaculture, livestock, waterways) in Germany between 2000-2022. We combine several dimensionality reduction algorithms (PCA, ISOmap, self-organizing maps) to generate robust and interpretable representations of the drought impacts. Our results show characteristic patterns for both particular drought events and regions. Also, the applied methods provide a low-dimensional representation of the multivariate socio-economic drought impacts. This research provides a methodological contribution to the holistic, empirical investigation of co-occurring drought impacts. The proposed methods can inform risk models, and policy-makers on the urgency of cross-sectoral governance approaches. Also, the proposed method could apply to other hazards or compound events.

How to cite: Sodoge, J., Kuhlicke, C., Mahecha, M., and de Brito, M.: Drought impact profiles: Analyzing multivariate socio-economic drought impacts using nonlinear dimensionality reduction, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-286, https://doi.org/10.5194/egusphere-egu23-286, 2023.

16:30–16:40
|
EGU23-14314
|
NH9.2
|
ECS
|
On-site presentation
Domenico Bovienzo, Sepehr Marzi, Letizia Monteleone, Jaroslav Mysiak, and Jeremy Pal

Climate change is projected to increase the frequency and intensity of future droughts particularly affecting the most low-income countries directly dependent on local rainfed food security and livelihoods. Drought risk and its related impacts depend on the drought hazard, the exposure and the vulnerability of the different socioeconomic sectors and/or ecosystems as well as the adaptive capacity of affected locations. The Horn of Africa, which includes Ethiopia, is currently experiencing one of the most severe droughts in the last 40 years. This study applies a storyline approach to investigate changes in drought risk for Ethiopia combining vulnerability, hazard and adaptive capacity information for current and future projected climatic and socio-economic conditions using a subnational level composite indicator. For our analysis, we define drought based on the Standardised Precipitation-Evapotranspiration Index (SPEI) which characterises the deficits in local water availability based on the precipitation and potential evapotranspiration. SPEI is computed using bias corrected Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) project based on the Coupled Model Intercomparison Project Phase 6 (CMIP6). The Drought vulnerability assessment is carried out combining exposure, adaptive capacity and sensitivity indicators, using INFORM index developed by the Joint Research Centre of the European Commission to support humanitarian crisis and disaster decision-making. The analysis shows that future drought will increase people in need of food assistance both under current population and future population projections. If humanitarian aid and assistance are maintained at recent historical levels, these findings show a substantial increase in the required amounts. These conditions are exasperated when humanitarian access is impeded by local conditions such as the current conflict in Ethiopia, when imports are reduced by crises such as those associated with the Russian invasion of the Ukraine, and by pandemics such as COVID-19. Climate change mitigation is shown to reduce the vulnerability of Ethiopia through a reduction in drought hazard frequency and intensity. The framework presented in this study can be used as a policymaking tool to provide information on how to better prioritize future loss and damage funds and adaptation and mitigation investments to reduce population vulnerability and exposure.

How to cite: Bovienzo, D., Marzi, S., Monteleone, L., Mysiak, J., and Pal, J.: Current and future evolution of drought risk in Ethiopia: A framework to inform disaster risk reduction and climate change adaptation policies, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14314, https://doi.org/10.5194/egusphere-egu23-14314, 2023.

16:40–16:50
|
EGU23-5148
|
NH9.2
|
On-site presentation
Sungju Han, Torsten Masson, Sabrina Köhler, and Christian Kuhlicke

Individual adaptation is essential for achieving community resilience as well as coping with residual risks that have not been addressed by current structural schemes for reducing flood risks. At the same time, it also implies that individuals should have the resources and capacity to protect themselves. So far, this has been interpreted in the social vulnerability concept as accounting only for income, wealth, or other materially relevant factors, showing how much vulnerable people are exposed to more risk. However, individual behavioural adaptability has hardly been included in the current vulnerability assessment.

In light of this, this study proposes a novel way to expand and link social classes using well-established social vulnerability indicators (i.e. income, education, and job status) with socio-psychological and lifestyle elements theoretically and empirically known to influence individual protective behaviour. We conducted a bias-adjusted three-step Latent Class Analysis (LCA) with covariates (socio-psychological and lifestyle elements) and distal outcomes (adaptive behaviour). A household survey (n = 1,753) conducted between June and July 2020 in 11 cities in Saxony, Germany, was used.

The preliminary result shows that socio-psychological and cultural factors that influence individual decision-making on proactive adaptive behaviour co-vary with social classes based on their resource endowment. It also revealed that the lower class tends to have less implementation of costly adaptation methods, for example, structural measures on housing, while less costly measures did not make a significant difference. As a result, we recommend that, in addition to the lack of material endowment, which can be associated with an increased risk of exposure, individual inaction of protective behaviour motivated by socio-psychological traits be considered for social vulnerability.

How to cite: Han, S., Masson, T., Köhler, S., and Kuhlicke, C.: Bridge the gap: Linking social vulnerability and adaptive behaviour, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5148, https://doi.org/10.5194/egusphere-egu23-5148, 2023.

16:50–17:00
|
EGU23-9923
|
NH9.2
|
On-site presentation
Matthew Preisser, Paola Passalacqua, and R. Patrick Bixler

Despite major advancements in climate modeling, weather forecasting, and emergency preparedness, deadly floods continue to have a global reach, impacting Eastern Kentucky, USA (July 2022), Assam, India (2022), Cape Town, South Africa (2022), and Insul, Germany (July 2021) to name just a few. The goal of this work is to quantify and forecast in near-real time a flood’s impact at high spatial resolution by estimating how a household’s accessibility to critical infrastructure changes during and immediately after a storm. Our approach consists of a static transportation assignment cost function that solves for the user equilibrium traffic solution. By overlaying the road network with a near-real-time pluvial and fluvial inundation estimate, we estimate the degree to which flooding impacts households’ likely travel patterns to critical resources. The output consists of demand information on both the road and resource infrastructure networks, which we translate into resiliency and redundancy metrics. Our goal for this model is for it to be able to be rapidly deployed across the USA and potentially abroad to better serve communities who would otherwise not have access to such research and information tools. We present a case-study for Austin, Texas as a proof of concept and to highlight the critical decision-making information our approach can provide to those who need it most including emergency responders, flood managers, and residents themselves. Through this network approach, we can estimate who loses access to critical resources completely, whose access has diminished, how resource distribution is or isn’t equitable, hot spot nodes to prioritize remediation, and more. Our approach uses only open-source information including infrastructure, Earth observation, and point measurement data in our multilayer network. This data requirement allows our model to potentially be applicable in numerous regions across the globe. Our future work will explore using the network insights from this model in a dynamic model of adaptive capacity and human infrastructure. This will provide further insights on socio-hydrological interactions and how varying emergency response policies, government interventions, and human trends might impact the recovery trajectories of different communities.

How to cite: Preisser, M., Passalacqua, P., and Bixler, R. P.: A network-based disaster resilience metric for estimating individuals’ loss of access to critical resources during flooding, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9923, https://doi.org/10.5194/egusphere-egu23-9923, 2023.

17:00–17:10
|
EGU23-3422
|
NH9.2
|
On-site presentation
Anna Lea Eggert, Karsten Arnbjerg-Nielsen, and Roland Löwe

Human activities have a profound impact on climate and hydrological processes, contributing to changes in the frequency and severity of hydrological extremes and, consequently, growing socioeconomic vulnerability [1]. Rising sea levels, continuous urban development in low-lying coastal areas, and corresponding changes in flood risk have resulted in devastating flood impacts. Different Flood Risk Management (FRM) strategies have been adopted in various socioeconomic contexts and spatiotemporal scales, the most prevalent being structural protection. In recent years, numerous scholars have raised concerns about this approach, as studies have shown that increasing protection levels can increase socioeconomic vulnerabilities e.g., [2]. FRM strategies alter the dynamics of risk manifested in sociohydrological systems, which must be disentangled to avoid unintended consequences.
In the “Cities and rising sea levels” project, scientists from different research disciplines, including hydrology, architecture, landscape architecture, and urban planning, collaborate to tackle these challenges. Combining multidisciplinary knowledge has been central to exploring the cross-sectoral processes involved in FRM. In the present study, we focused on (1) uncovering the cascading effects, including unintended consequences of FRM, as well as (2) highlighting the potentials for holistic assessments of FRM strategies.
Our methods include the development of a Causal Loop Diagram (CLD) model describing critical sociohydrological processes of coastal cities operating at different spatial and temporal scales. We identified dynamic feedbacks between (1) flood risk, urban development and economic wealth, (2) flood risk, urban development and social equity, and (3) flood risk, trust in authorities, and institutional capacity, among others. . Based on the CLD, we analyzed key feedback mechanisms and their manifestation in theory and practice. Further, we explored the impacts of different FRM strategies on these feedback mechanisms to uncover differences in impacts on socioeconomic vulnerabilities and wider cross-sectoral impacts. The presentation will present and explore the conceptual model through semiquantitative analyses (Fuzzy Cognitive Maps (FCMs)) and spatiotemporal assessments using a specific case study. We aim at (1) getting case-specific insights into the dynamics produced by the local interplay of flooding events and socioeconomic processes influencing vulnerabilities, and (2) suggesting pathways for new integrated ways of FRM.

References
[1] IPCC, Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. In Press., 2022.
[2] R. W. Kates, C. E. Colten, S. Laska, and S. P. Leatherman, “Reconstruction of New Orleans after Hurricane Katrina: A research perspective,” Proc. Natl. Acad. Sci. U. S. A., vol. 103, no. 40, pp. 14653–14660, Oct. 2006, doi: 10.1073/PNAS.0605726103/ASSET/C486E9DB-5923-43C0-9881-2B57734F2A7C/ASSETS/GRAPHIC/ZPQ0410637570002.JPEG.

How to cite: Eggert, A. L., Arnbjerg-Nielsen, K., and Löwe, R.: Feedbacks between City Development and Coastal Flood Risk Management: A Systems Thinking Approach, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3422, https://doi.org/10.5194/egusphere-egu23-3422, 2023.

17:10–17:20
|
EGU23-16481
|
NH9.2
|
ECS
|
On-site presentation
Isabel Hagen, Sanne Schnyder, Inés Yanac León, Sirkku Juhola, Veruska Muccione, and Christian Huggel

The highly populated Peruvian Andes is impacted by a multitude of climate-related risks. Comprehensive climate risk management and adaptation measures can bring risks down to an acceptable level, as determined by the local population. However, increased magnitude and frequency of risks, together with the possibility of reaching adaptation limits, are hindering risk reduction. Adaptation limits are reached due to a complex interplay between socio-economic, cultural, political, institutional, technical and bio-physical factors. Whilst there is an emerging conceptual understanding of adaptation limits, there is little empirical research investigating limits in real-world settings.

The aim of this study is to identify and define the limits of adaptation on a local scale, which limits are approaching and which have already been reached. We investigate the limits of adaptation in two catchments in the Peruvian Andes. The most prevalent climate-related risks in these two regions are from glacial lake outburst floods, landslides, shifts in precipitation patterns, and glacier retreat. We use a conceptual framework developed by Juhola et al. (unpublished), and determine adaptation limits and the intolerable risks space through investigating human wellbeing, governance systems, ecosystem functions and climate hazards in the two localities. The data was collected through a thorough literature review, together with 50 semi-structured interviews conducted in May-July 2022; 28 with local residents in the Río Santa and Salkantay catchments, and 22 interviews with experts from 14 different local and national institutions and NGOs. The interviews were analysed in Atlas.ti using a content analysis approach. We emphasize the focus on basic needs and wellbeing, to encompass not only what are obvious losses from climate impacts, such as loss of life or livelihood, but also more intangible losses, such as limited mobility, loss of a social network, or loss of local knowledge. The conclusions of this study can help decision makers and practitioners improve the positive impact of future risk management and adaptation projects in the two regions.

How to cite: Hagen, I., Schnyder, S., Yanac León, I., Juhola, S., Muccione, V., and Huggel, C.: Limits of adaptation to climate-related risks in the Peruvian Andes: A case study in the Río Santa and Salkantay catchments, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16481, https://doi.org/10.5194/egusphere-egu23-16481, 2023.

17:20–17:30
|
EGU23-17449
|
NH9.2
|
On-site presentation
Iain Willis and Myrto Papaspiliou

Societal resilience is built upon effective risk transfer strategies. For most developed countries in the world, insurance and reinsurance continues to be the most effective method of sharing this burden and reducing the need for state intervention. However, it’s becoming increasingly clear that the probabilistic (CAT) models used to price natural hazard risk are struggling to capture the increasingly dynamic changes of the climate and our level of interconnection.

The Gallagher Research Centre (GRC) was established recently to support reinsurance stakeholders navigate an increasingly complex risk management landscape. Though probabilistic and deterministic natural catastrophe models were first pioneered in the early 1960’s (Friedman, 1984) it wasn’t until the 1990’s, and the combined losses from Hurricane Andrew ($27.3 billion USD) and the Northridge Earthquake ($25 billion USD) that such models began to be fully embraced by the mainstream reinsurance industry (Reinsurance News, 2023).

While significant and continued progress has been made in the precision and scalability of these models in the last 30 years, climate change and an increasingly globalized world mean the relative impacts of natural hazards are becoming far more complex and diverse than most models successfully capture. This leads to an increasing basis risk and potentially less utility of the models. This session will outline the growing research concerns of focus for the GRC, including how can stochastic models built around historical periods truly capture the non-stationarity of risk we see occurring for wind and flood perils? Should models capture the seasonal dependencies between perils to more accurately price aggregate insurance risk? Should future model development focus on the compounded scenarios? 

 

Friedman, D. G. (1984). Natural hazard risk assessment for an insurance program. Geneva Papers on Risk and Insurance, 57-128.

Reinsurance News (2023). Last accessed 10/01/2023. https://www.reinsurancene.ws/insurance-industry-losses-events-data/

How to cite: Willis, I. and Papaspiliou, M.: The urgency for (re)insurance probabilistic (CAT) models to capture the dynamics of an increasingly interconnected world, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17449, https://doi.org/10.5194/egusphere-egu23-17449, 2023.

17:30–17:40
|
EGU23-4821
|
NH9.2
|
ECS
|
On-site presentation
Jeongha Lee and Seokhwan Hwang

As industrialization and urbanization progress around the world, more complex and large-scale complex disasters are occurring, causing numerous casualties and property damage every year. As climate change gradually accelerates and its impact grows, such as recent cold waves and heavy snow in the United States and abnormal temperatures in Europe, it is difficult to predict with existing physical modeling alone. Recently, disasters are gradually expanding in the form of covering not only natural disasters but also various social disasters. Social disasters cover disasters such as fires, infectious diseases, and fine dust caused by human activities. Unlike natural disasters, it is difficult to measure numerical values and predict occurrence patterns in real time, so it is very important to respond quickly through information sharing. There is a limit to establishing the same response system globally to respond to disasters that may occur worldwide, so it is necessary to develop a platform that can quickly share cases while being economical. With the recent development of communication technology, about 70% of the world's population uses smartphones, and various unstructured data are being generated in real time through various social media channels. Individuals act as a sensor and can share their location or current situation in real time. Therefore, the purpose of this study is to develop crowd sourcing technology using social media, analyze the collected data, and present ways to use it in the event of a disaster. In this study, a platform was established to collect and analyze disaster-related SNS data such as floods, fine dust, and forest fires, and it was designed so that users could receive information through websites and apps. As a result of application to various disaster cases in Korea, the temporal and spatial correlation between disaster occurrence patterns and social media data was high, and the possibility of using initial monitoring methods was proved. This result can be applied to all disaster disasters or crimes, and it is expected to be highly useful as it can quickly verify disaster thoughts and share cases in real time.

 

How to cite: Lee, J. and Hwang, S.: A Study on the Monitoring of Complex Disaster Using Crowd Source Data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4821, https://doi.org/10.5194/egusphere-egu23-4821, 2023.

17:40–17:50
|
EGU23-567
|
NH9.2
|
ECS
|
On-site presentation
Silvia De Angeli, Stefano Terzi, Davide Miozzo, Lorenzo Stefano Massucchielli, Joerg Szarzynski, Fabio Carturan, and Giorgio Boni

The Disaster Risk Management Cycle (DRMC) is a common reference for the international Disaster Risk Management (DRM) community to describe the management of catastrophic anthropogenic and natural events worldwide. Implementing this approach, disaster management is described by a series of separate and consecutive phases (e.g., preparedness, response, and recovery). However, the current DRMC is not able to successfully cover the dynamics of multi-hazard risk scenarios, particularly those involving both sudden- (e.g., earthquakes or flash floods) and slow-onset hazards (e.g., pandemics or  droughts).

Starting from such a complex scenario we propose a ‘parallel phases’ DRM model accounting for the management of interacting sudden- and slow-onset hazards. The framed ‘parallel phases’ model allows to overcome the limitations of the existing models when dealing with complex multi-hazard risk conditions. We supported the identified limitations analysing Italian Red Cross data dealing with past and ongoing emergencies including the COVID-19 pandemic. Key findings from the analysis involve: (i) the spatial-temporal differences between sudden-onset events and pandemic disaster management; (ii) the high demand for emergency response resources during pandemics in comparison to other emergencies; (iii) the need for the DRM system to adjust the response to cope with the pandemic seasonality; (iv) the system over-exposure to pandemic response activities reducing the number of resources for preparedness and entering the system into an unpreparedness negative loop.

Overall, the combination of the key findings that emerged from the management of the COVID-19 pandemic in Italy brought out three main guidelines for advancing multi-hazard DRM by applying our ‘parallel phases’ model:

  • Managing the system with parallel phases. A ‘parallel phases’ DRM allows the system to exploit the low emergency intensity of the slow-onset hazards seasonality for preparedness actions while also preparing for any other hazard that can have relevant impacts on the system. Such an approach allows the DRM system to escape from an unpreparedness negative loop. 
  • Keeping the DRM system capacity far from depletion. The DRM system can learn how to efficiently deploy the available resources keeping its capacity far from total depletion. If the DRM system is able to save part of its capacity, it can continue with the increase of internal resources while also making them available for international mutual support in case of multi-hazard risk. Such a condition triggers a positive loop in the increase of the DRM capacity.
  • Impact-based forecasting for multi-hazard disaster risk management. The implementation of multi-hazard seasonal impact-based forecasts fosters the planning of appropriate anticipatory actions, combining the prediction of slow-onsets waves with the seasonality of sudden-onsets.

Overall, the proposed ‘parallel phases’ model is able to capture the complex management dynamics to deal with the increasingly frequent slow-onset and multi-hazard events, introducing a change of perspective from the cyclic, consecutive-phases, and single-hazard DRM approach. For this reason, the ‘parallel phases’ model can strengthen and boost current and future international policies on multi-hazard DRM towards an effective implementation at a local scale.

How to cite: De Angeli, S., Terzi, S., Miozzo, D., Massucchielli, L. S., Szarzynski, J., Carturan, F., and Boni, G.: Managing the co-occurrence of natural hazards and pandemics with a new parallel phases DRM model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-567, https://doi.org/10.5194/egusphere-egu23-567, 2023.

17:50–18:00

Posters on site: Mon, 24 Apr, 10:45–12:30 | Hall X4

Chairpersons: Korbinian Breinl, Johanna Mård, Giuliano Di Baldassarre
X4.36
|
EGU23-6810
|
NH9.2
|
ECS
Riccardo Biella, Sina Khatami, Luigia Brandimarte, Maurizio Mazzoleni, and Giuliano Di Baldassarre

Climate services are expected to deliver better climate adaptation by providing decision-makers with timely, salient, credible, legitimate, and accessible climate information. Nonetheless, climate services’ impact on long-term adaptation remains poorly understood due to their ambiguous protocols, quality standards, and inadequate monitoring and evaluation processes.

The aim of this study is to present the underpinnings of a framework representing the causal mechanisms and feedback interactions between adaptation to hydrometeorological extremes, i.e. floods and droughts, and climate services among the partner living labs of the I-CISK project (https://icisk.eu). To this end, a qualitative investigation based on interviews and surveys of the living labs’ stakeholders is performed. Following, the findings from the qualitative analysis are iteratively discussed with the stakeholders and presented as a causal loop diagram, highlighting feedback loops in the coupled human-climate system. Finally, the emerging dynamics are described using system archetypes.

This research offers a systemic tool for evaluating the long-term dynamics of adaptation to hydrometeorological extremes while building the bases for further research in the living labs. Moreover, it shows the efficacy of system dynamics tools for informing adaptive policy-making.

How to cite: Biella, R., Khatami, S., Brandimarte, L., Mazzoleni, M., and Di Baldassarre, G.: Conceptualizing the long-term interactions between climate services and adaptation to hydrometeorological extremes, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6810, https://doi.org/10.5194/egusphere-egu23-6810, 2023.

X4.37
|
EGU23-10110
|
NH9.2
Frank Davenport, Shrad Shukla, Donghoon Lee, Patrese Anderson, Greg Husak, and Chris Funk

The potential for predictive models based on earth observations (EO) and survey data to assist in famine early warning and other development applications is rapidly growing. However, while the spatial-temporal extent of EO data is complete, high quality survey data is generally limited in spatial and temporal scope. The perennial question in all predictive analysis, and especially when trying to move from research to operational application in the developing world is: If we create a forecast model from region A (based on observed outcomes) can we apply the same model in region B, where we do not observe or have limited observations of those outcomes? Prior research has proposed examining the Area of Influence (AoI) based on structurally similar characteristics in the EO predictors. We expand on and evaluate this approach in the context of grain yield forecasting in Sub-Saharan Africa (SSA). Specifically, we evaluate an AoI methodology established for generating raster surfaces and apply it to vector supported grain data.  We ask the following questions: What are the key characteristics that make a forecast fit for one country work in another country? Can pooling models across multiple countries provide more accurate out-of-sample estimates than a model fit to one country or district? Does AoI change through the season? Does a model fit for in early season have the same AoI as a model fit late in the season.

 

How to cite: Davenport, F., Shukla, S., Lee, D., Anderson, P., Husak, G., and Funk, C.: Testing Spatial Out-of-Sample Area of Influence for Grain Forecasting Models:  How does out of Spatial Out-of-Sample AoI Change through the Season? , EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10110, https://doi.org/10.5194/egusphere-egu23-10110, 2023.

X4.38
|
EGU23-10716
|
NH9.2
Urban Growth of Houston, Texas over the Last Two Decades: Implications for Heat Stress
(withdrawn)
Udaysankar Nair, Andrew Blackford, Joey Bonucchi, Brian Freitag, Manil Maskey, and Aaron Kaulfus
X4.39
|
EGU23-5611
|
NH9.2
|
ECS
Jiting Tang and Saini Yang

Studying the spatiotemporal patterns of urban road traffic under extreme weather is a key step to building a climate-resilient city. Although existing researches model and simulate traffic states from different perspectives, the traffic forecasting of the urban road networks under extreme weather is seldom addressed. In this paper, a novel Knowledge-driven Attribute-augmented Attention Spatiotemporal Graph Convolutional Network framework is proposed to predict urban road traffic under wind and rain especially in tropical cyclone disasters. Considering the disaster conditions, we model the external dynamic hazard attributes and static environment attributes, and designed an attribute-augmented unit to encode and integrate these factors into the deep learning model. The model is combined with the graph convolutional network (GCN), the gated recurrent unit (GRU), and the attention mechanism. Experiments demonstrate that the predictability of traffic speed can be greatly increased by supplementing the disaster-related factors, the prediction accuracy reaches 0.79. The proposed approach outperforms baselines by 12.16%-31.67% on real-world Shenzhen’s traffic datasets. The model also performs robustly on different road vulnerabilities and hazard intensities. The model errors are mainly occurred in the early peak with extreme wind and rain and the coastal area in the southeast of Shenzhen because of the greater uncertainty. The framework and findings provide a valuable reference for the decision-making of traffic management and control prior to a disaster to alleviate traffic congestion and reduce the negative impact of disasters.

How to cite: Tang, J. and Yang, S.: A traffic prediction framework under extreme weather combined disaster knowledge and deep learning, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5611, https://doi.org/10.5194/egusphere-egu23-5611, 2023.

X4.40
|
EGU23-6850
|
NH9.2
|
ECS
Cross-geography adaptation to climate change in coffee-growing regions: is it risk-specific or generalized?
(withdrawn)
Gina Maskell, Sarah Murabula Achola, Sabine Undorf, Paula Romanovska, and Christoph Gornott
X4.41
|
EGU23-7967
|
NH9.2
Olga Nardini, Stefano Morelli, Veronica Pazzi, and Sara Bonati

Social media have the potential to significantly influence the disaster risk understanding of natural events of climatic and geological origin, e.g., earthquakes, volcanic eruptions and landslides. Given their considerable diffusion, nowadays they represent a valid support during emergency management processes thanks to their multiple uses in all the different phases of the disaster cycle. The presented results have been achieved carrying out a literature review in the framework of the European H2020 project LINKS ('Strengthening links between technologies and society for European disaster resilience') which aims to strengthen the link between technology and society to improve resilience in four European countries associated with five different risk scenarios. The aim of this research was to investigate how social media influence and impact vulnerability and risk perception and how the increased use of social media as a communication tool during a disaster is shaped by the way the two concepts interact and are conceptualised. The main results are that through social media, it is possible to raise people's awareness of the disaster, also by working on each individual's trust in those who provide information, but also to disseminate useful information and alerts to the population to keep abreast of real-time events, to connect citizens with each other in order to reduce distances and provide psychological support, and to create a social network for those in need. Additionally, social media can be used to manage an emergency and coordinate volunteer actions. The concepts of vulnerability and risk perception are extremely important to be considered when talking about geological hazards and disasters. They are two interconnected concepts that need to be pursued hand in hand in emergency management. The main challenges and factors impacting the use of social media concern access, quality and reliability of information, trust, and awareness of the news being provided, but also personal experience and geographical, social and demographic factors that may influence the way information is perceived and understood. The perception of geological risks directly influences people's preparedness and the way they act, helping anyone to understand the scope of the event and the potential risks that could occur, in order to make informed decisions on how to react. Furthermore, a real understanding of vulnerability influences the resilience of local communities in relation to disasters and can in turn be influenced using social media. Social media can also amplify public fear and concern about the disaster, especially if there is a lot of misinformation or sensationalism about the event. This can lead to an overestimation of risks and an increased sense of vulnerability among the population. These results could be helpful in identifying possible methods and approaches to study these issues in the future.  

How to cite: Nardini, O., Morelli, S., Pazzi, V., and Bonati, S.: Social media, vulnerability, and risk perception: three main points for geological disaster management, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7967, https://doi.org/10.5194/egusphere-egu23-7967, 2023.

X4.42
|
EGU23-3031
|
NH9.2
Seung-Hee Oh, Hyunjoo Kang, and Sang-Lim Ju

Between 2020 and 2022, when South Korea experienced Covid-19, it suffered from multiple natural disasters, including typhoons, forest fires, and earthquakes, as well as infectious diseases. Recently, not only in Korea but also worldwide due to climate change, the number and scale of natural disasters are increasing every year, and the damage caused by them is becoming more and more serious. We analyzed big data on disasters in South Korea to identify trends in disasters caused by climate change. So, between 2012 and 2022, we downloaded over 100,000 open data on emergency disaster alert messages (by mobile network Cell Broadcasting Service) provided by the central government and local governments to the general public through Public data portal (https://www.data.go.kr/) open API(Application Programming Interface). And we visualized the collected raw big data based on GIS after refinement, classification((Natural and social disasters, disaster type, disaster level, CBS msg type, emergency disaster message sending agency, etc.), and subdivision by city (we call it Si, Gun, Gu) unit area. Then, it was displayed based on GIS according to the type of disaster. We performed visualization work to derive the results of climate change trends in South Korea by disaster type and by region(Si, Gun, Gu).
Through this, it was possible to identify the types of disasters that are becoming more severe in South Korea according to climate change. Also, based on these results, we were able to identify which disasters each region would be vulnerable to. In addition, based on these results, we were able to identify which disasters are particularly vulnerable according to the characteristics of each region and which disasters it is best to strengthen preparation for in the future.
The results of analyzing the past history big data of our emergency disaster messages can be usefully used to present preventive and prepared plans for future disasters by central and local governments.
This research was supported by a grant (20008820) of Disaster-Safety Inter-Miniterial Cooperation Program funded by Ministry of Interior and Safety (MOIS, Korea)

How to cite: Oh, S.-H., Kang, H., and Ju, S.-L.: Analysis of natural disaster vulnerability by region through the use of big data of emergency disaster message history, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3031, https://doi.org/10.5194/egusphere-egu23-3031, 2023.

X4.43
|
EGU23-16045
|
NH9.2
Carlo De Michele, Fabiola Banfi, and Viola Meroni

Compound climate-related (or weather-related) events are complex events characterized by the interactions between various physical processes across multiple spatial and temporal scales, generated by meteorological variables, and provoking extreme impacts. Compound climate-related events often include the joint occurrence of multi-hazards like landslides and floods, or heatwaves, droughts and wildfires.

In literature, databases of natural hazards are in general single hazard, like databases of floods (European Flood Database, AVI database), landslides (Global Fatal Landslide Database , AVI database), droughts (European Drought Observatory).

The assessment and understanding of compound events requires an integrated perspective, with the integration of data from multiple variables, combining multiple databases.

In this presentation, we try to address this emerging need, illustrating a possibility of building a compound events database, and presenting some examples.

How to cite: De Michele, C., Banfi, F., and Meroni, V.: Compound events from the databases, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16045, https://doi.org/10.5194/egusphere-egu23-16045, 2023.

Posters virtual: Mon, 24 Apr, 10:45–12:30 | vHall NH

Chairpersons: Johanna Mård, Giuliano Di Baldassarre, Korbinian Breinl
vNH.6
|
EGU23-2266
|
NH9.2
|
ECS
Similar but different? Climate risk archetypes can help the municipal adaption process
(withdrawn)
Nils Riach and Rüdiger Glaser